A Single-Site Mutation at Ser146 Expands the ... - ACS Publications

Aug 19, 2016 - Institute for Innovative Learning, Mahidol University, Nakhon ... for Innovation in Chemistry, Faculty of Science, Mahidol University, ...
1 downloads 0 Views 1MB Size
Articles pubs.acs.org/acschemicalbiology

A Single-Site Mutation at Ser146 Expands the Reactivity of the Oxygenase Component of p‑Hydroxyphenylacetate 3‑Hydroxylase Taweesak Dhammaraj,† Chatchadaporn Pinthong,†,‡ Surawit Visitsatthawong,§ Chanakan Tongsook,† Panida Surawatanawong,§ and Pimchai Chaiyen*,† †

Department of Biochemistry and Center for Excellence in Protein and Enzyme Technology, Faculty of Science, Mahidol University, Rama 6 Road, Bangkok 10400, Thailand ‡ Institute for Innovative Learning, Mahidol University, Nakhon Pathom 73170, Thailand § Department of Chemistry and Center of Excellence for Innovation in Chemistry, Faculty of Science, Mahidol University, Rama 6 Road, Bangkok 10400, Thailand S Supporting Information *

ABSTRACT: The oxygenase component (C2) of p-hydroxyphenylacetate (4-HPA) 3-hydroxylase (HPAH) from Acinetobacter baumannii catalyzes the hydroxylation of various phenolic acids. In this report, we found that substitution of a residue close to the phenolic group binding site to yield the S146A variant resulted in an enzyme that is more effective than the wild-type in catalyzing the hydroxylation of 4-aminophenylacetate (4-APA). Product yields for both wild-type and S146A enzymes are better at lower pH values. Multiple turnover reactions of the wild-type and S146A enzymes indicate that both enzymes first hydroxylate 3-APA to give 3-hydroxy-4-aminophenylacetate (3-OH-4-APA), which is further hydroxylated to give 3,5-dihydroxy-4aminophenylacetate, similar to the reaction of C2 with 4-HPA. Stopped-flow experiments showed that 4-APA can only bind to the wild-type enzyme at pH 6.0 and not at pH 9.0, while it can bind to S146A under both pH conditions. Rapid-quench flow results indicate that the wild-type enzyme has low reactivity toward 4-APA hydroxylation, with a hydroxylation rate constant (kOH) for 4-APA of 0.028 s−1 compared to 17 s−1 for 4-HPA, the native substrate. In contrast, for S146A, the hydroxylation rate constants for both substrates are very similar (2.6 s−1 for 4-HPA versus 2.5 s−1 for 4-APA). These data indicate that Ser146 is a key catalytic residue involved in optimizing C2 reactivity toward a phenolic compound. Removing this hydroxyl group expands C2 activity toward a non-natural aniline substrate. This understanding should be helpful for future rational engineering of other twocomponent flavin-dependent monooxygenases that have this conserved Ser residue. Most of the flavin-dependent monooxygenases that can hydroxylate aromatic compounds require the presence of a phenolic group on the substrate to facilitate the electrophilic aromatic substitution reaction.5,9,10 In addition to phenolic compounds, a few enzymes can hydroxylate aniline compounds such as the FADH2-utilizing monooxygenase (GTNG_3160) encoded in the anthranilate degradation gene cluster in Geobacillus thermodenitrif icans NG80-2 which catalyzes the conversion of anthranilate (or 2-aminobenzoate) into 3hydroxyanthranilate.11 Some hydroxylases that can use aniline derivatives such as 4-aminophenylalkanoic acid or 4-aminobenzoic acid as substrates have combined activities of hydroxylation with either deamination or decarboxylation.12−14 p-Hydroxybenzoate hydroxylase from P. fluorescens15 and P. aeruginosa16 can catalyze ortho-hydroxylation of aniline

F

lavin-monooxygenases are important biocatalysts.1,2 These enzymes can catalyze the regio-specific incorporation of a single oxygen atom in the form of a hydroxyl or epoxide group into a wide range of substrates. Examples of enzymes in this class that are useful in biocatalysis include Baeyer−Villiger monooxygenases, styrene monooxygenase, and various aromatic hydroxylases.3−5 As biocatalysts are instrumental for the development of green and sustainable manufacturing processes, the engineering of biocatalysts to broaden their repertoire of natural enzymatic activities has the potential to provide a comprehensive enzymatic toolbox to fit a variety of applications.6 The expansion of enzymatic specificity to encompass reactivity toward non-native substrates is generally accomplished by engineering enzymes through a random, semirandom, or rational approach.6−8 The rational engineering approach, which is the most cost-effective method, requires an understanding of the catalytic reaction mechanism and of how the reaction reactivity and specificity is controlled by active site residues. © XXXX American Chemical Society

Received: May 9, 2016 Accepted: August 19, 2016

A

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX

Articles

ACS Chemical Biology

from the coupled enzymes (if multiple turnover reactions were employed). Under these conditions, the enzyme only catalyzed one catalytic cycle. Because the complex of C2 and reduced FMN (C2:FMNH−) was the limiting reagent, it was oxidized in the excess of oxygen and could not be regenerated in the absence of reducing agent. The results indicated that 4mercaptophenylacetate, 4-methoxyphenylacetate, 4-nitrophenylacetate, 4-fluorophenylacetate, and 4-chlorophenylacetate could not be used as substrates (data not shown) and only 4APA could be used (Table 1). HPLC/DAD/MSD chromato-

derivatives (p-aminobenzoate) but with much smaller hydroxylation rate constants (more than 10-fold) compared to those for the native phenolic substrates. The hydroxylation of 4aminophenylalkanoic acid to produce the corresponding 3hydroxy-4-aminophenylalkanoic acid is potentially useful in biocatalysis because many of these compounds, such as 3hydroxy-4-aminophenylacetic acid (3-OH-4-APA), are precursors for the synthesis of pharmacologically active compounds.17−22 p-Hydroxyphenylacetate 3-hydroxylase (HPAH) from Acinetobactor baumannii makes a good candidate starting enzyme for exploratory engineering to produce an enzyme that can prepare 3-hydroxy-4-aminophenylacetate (3-OH-4-APA) from 4-aminophenylacetate (4-APA). Native HPAH catalyzes orthohydroxylation of p-hydroxyphenylacetate (4-HPA) to form 3,4-dihydroxyphenylacetate (3,4-DHPA). The enzyme is a twocomponent flavin-dependent monooxygenase that consists of reductase (C1) and oxygenase (C2) components.23,24 The reaction mechanism of C2 is one of the best understood among the two-component flavin-dependent monooxygenases,25−32 and X-ray crystal structures are available.33 In addition to 4HPA, C2 can also catalyze the ortho-hydroxylation of other phenolic acids with R groups such as −COOH, −CH2COOH, −CH2CH2COOH, and trans-CHCHCOOH.34 C2 has many properties that make it suitable for biocatalytic applications including having a high yield of overexpression and exhibiting good stability over a wide-range of pH conditions (pH 6−10)28 and temperatures (25−40 °C for more than 24 h).34 Therefore, expansion of the activity of C2 to go beyond hydroxylation of phenolic acids will make this flavin-monooxygenase a versatile enzyme useful for inclusion in a biocatalytic toolbox for multiple applications. In this report, we found that engineering of C2 at Ser146 yielded the S146A variant that is much more effective than the wild-type enzyme in catalyzing ortho-hydroxylation of 4aminophenylacetic acid (4-APA) to generate 3-hydroxy-4aminophenylacetic acid (3-OH-4-APA). The kinetics of ligand binding to wild-type and S146A toward 4-APA and 4-HPA were shown to be different. Measurement of the hydroxylation rate constants of the wild-type and S146A enzymes revealed that, while the wild-type enzyme has a much greater reactivity toward 4-HPA, the reactivity of S146A toward 4-APA and 4HPA are not different. Overall, we have shown that a single-site mutation at S146 can expand the activity of C2 to go beyond the natural phenolic acid substrate to include an aniline compound.

Table 1. Product Formation from Single Turnover Reactionsa of Wild-Type C2, Ser146, and His120 Variants with 4-APA (0.2 mM) at pH 6.0−9.0

C2 enzyme wild-type

S146A

S146C

H120N H120D H120E H120K control (no enzyme)

pH

% yield of 3-OH-4-APAb

6 7 8, 9 6 7 8 9 6 7 9 6 7, 9 6, 7, 9 6, 7, 9 6, 7, 9 6, 9

40 ± 1 31 ± 5 100.0 95 77 66 28.7 9

± ± ± ± ± ±

0.1 3 2 5 0.8 3

17.1 ± 0.1

a

Under this experimental condition, the enzyme could only perform one catalytic cycle because C2:FMNH− (limiting reagent) was all oxidized in the excess of oxygen. With no excess reducing agent, FMNH− could not be regenerated. The reactions were performed by mixing a 250 μL solution of C2 (200 μM):FMNH− (100 μM) in 10 mM NaH2PO4 at pH 7.0 with a 250 μL solution of 400 μM substrate in various aerobic buffers including 100 mM NaH2PO4 at pH 6.0 and pH 7.0, 100 mM Tris-HCl at pH 8.0, and 100 mM Gly-NaOH at pH 9.0. The reactions were left at RT for 15 min and then quenched with 500 μL of 0.15 M HCl. The quenched solutions were filtered through Microcon centrifugal units (10 kDa cutoff), and the filtrates obtained were analyzed using HPLC/DAD/MSD. bThe % yield of product was calculated from the amount of substrate consumed compared to the limiting agent (reduced flavin) in the reaction.



RESULTS AND DISCUSSION Rational engineering of C2 was performed in order to obtain a variant that is superior to the wild-type enzyme in catalyzing hydroxylation of 4-APA. Transient kinetics (stopped-flow and rapid-quench flow techniques) indicated that the hydroxyl group of Ser146 is key to specifying C2 reactivity toward a phenolic compound. Hydroxylation of 4-APA by wild-type and S146A is more efficient at lower pH. Density functional theory calculations of the binding energy indicate that ligand binding is more favorable when 4-APA or His120 is protonated. Single Turnover Reactions of Wild-type Enzyme with 4-APA at pH 7.0. Single turnover reactions of wild-type C2 with 4-APA and other aromatic compounds with different types of substituents were carried out at pH 7.0 (see Methods). We chose to use single turnover reactions to identify the hydroxylation product in order to avoid complicating factors

grams of the wild-type C2 reaction with 4-APA are shown in Supporting Information Figures S2−S4. The peak for the product as detected by HPLC/MSD-positive mode at 12.88 min showed an m/z of 168 (Supporting Information Figure S2d), and the peak for the substrate at 15.12 min showed an m/ z of 152 (Supporting Information Figure S2c). The data indicate that the product has an m/z value of 16 AUs greater than the substrate (m/z 152), suggesting that the product is the hydroxylated 4-APA. This product has the same HPLC and mass spec profiles as the standard 3-hydroxy-4-aminophenylacetic acid (3-OH-4-APA; Supporting Information Figure S5), indicating that the wild-type C2 catalyzes ortho-hydroxylation of B

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX

Articles

ACS Chemical Biology

enzyme (∼31 ± 5%). The chromatographic and MSD characteristics of the product from the S146A reaction with 4-APA are the same as those of the product from the wild-type enzyme (Supporting Information Figure S4a−d). Reactions of Wild-type C2 and His120 and Ser146 Variants with 4-APA at Various pH’s. Single turnover reactions of wild-type C2, S146A, S146C, H120E, H120K, and H120N with 4-APA (0.2 mM) were carried out over a pH range of 6.0 to 9.0 (see Methods), and the amount of product formed by each variant is shown in Table 1. H120D, H120E, and H120K were not able to catalyze the hydroxylation of 4APA at any of these pH values, while the others were more effective at the lower pH values. For most enzymes, the hydroxylation yield of 4-APA is greater at lower pH’s, and they cannot catalyze hydroxylation of 4-APA above pH 7.0. H120N hydroxylates 4-APA only at pH 6.0. Wild-type C2 and S146C hydroxylates 4-APA only at pH 6.0 and 7.0. S146A, the enzyme with the highest efficiency of 4-APA hydroxylation, was able to catalyze the hydroxylation of 4-APA over a wide pH range (6.0−9.0). Based on these results, S146A was chosen for further investigations by transient kinetics to compare with the wildtype C2. Multiple Turnover Reactions of Wild-Type C2 and S146A with 4-APA at pH 6.0. Multiple turnover reactions of wild-type C2 and S146A with 4-APA at pH 6.0 (Supporting Information Scheme S1a) were carried out to investigate the bioconversion of 4-APA by these enzymes on a larger scale. The results show that both enzymes first catalyze hydroxylation of 4-APA to form 3-OH-4-APA. The 3-OH-4-APA was then hydroxylated to form 3,5-dihydroxy-4-aminophenylacetic acid (3,5-diOH-4-APA; Supporting Information Figure S6, Tables S1 and S2, and Scheme S1b). Structures of the single and double hydroxylated products were identified using 1H and 13C NMR spectroscopy (Supporting Information Figures S7−S10). Results of the multiple turnover reactions also indicate that the reaction of S146A with 4-APA or 3-OH-4-APA proceeds much faster than the reaction of the wild-type enzyme. The abilities of wild type and S146A to catalyze double hydroxylation of 4-APA is similar to the reactions of wild-type C2 and Y398S with 4HPA and p-coumaric acid,34 in which the double hydroxylated product is the final product. Transient Kinetic Analysis of Wild-Type C2 and S146A Reactions to Investigate the Determining Factors that Make S146A More Efficient than the Wild-Type Enzyme in 4-APA Hydroxylation. The results in the sections above indicate that S146A is more efficient than the wild-type enzyme in the hydroxylation of 4-APA, especially at higher pH’s (Table 1). We carried out transient kinetic experiments to test what catalytic steps in the reaction were affected by the substitution. A. Reaction of Wild-Type C2 and S146A:FMNH− with Oxygen in the Absence of Substrate at pH 6.0 and pH 9.0 Observed by Stopped-Flow Spectrophotometry. In the absence of substrate, the reactions of C 2:FMNH − or S146A:FMNH− with oxygen can only result in the formation of C4a-hydroperoxyFMN and the subsequent elimination of H2O2 to give oxidized FMN (Scheme 1a).28,30 The rate constants of C4a-hydroperoxy-FMN formation were calculated from the first phase of the absorbance increase at 390 nm and at 446 nm, which indicate the formation of oxidized flavin (blue kinetic traces in Figures 2a,b and 3a,b, Supporting Figures S11a,b and S12a,b; Table 2). Overall, the rate constants associated with the two steps in both enzymes (Scheme 1a and

4-APA to form 3-OH-4-APA (product identity was also confirmed by NMR, discussed below). Table 1 indicates that only a 31 ± 5% yield of 3-OH-4-APA was generated from the single turnover reaction at pH 7.0. The control reaction in the absence of C2 did not show a peak at 12.88 min (Supporting Information Figure S3a and d). Reactions of His120 and Ser146 Variants with 4-APA. The results of the single turnover reaction above indicate that wild-type C2 is not an efficient biocatalyst for the hydroxylation of 4-APA. To improve the ability of C2 to hydroxylate 4-APA, the residues located close to the binding site of the phenolic moiety were subjected to enzyme engineering. On the basis of the X-ray structure of the C2:FMNH−:4-HPA complex (PDB code 2JBT),33 the hydroxyl group of 4-HPA is located close to His120 and Ser1462.8 Å (Figure 1a). A previous report30

Figure 1. (a) Active site structure of the C2:FMNH−:4-HPA complex (PDB code 2JBT).33 The optimized geometries of N-protonated 4APA and neutral His120 (b) and neutral 4-APA and protonated His120 (c) in the active site model in the form of C4a-hydroperoxyflavin. Bond distances are shown in Ångströms.

demonstrated that His120 and Ser146 are important for the hydroxylation of 4-HPA. His120 in particular may help in stabilizing the negative charge of the phenolate anion during the electrophilic aromatic substitution reaction. Mutation of Ser146 to Ala or Cys decreased the hydroxylation yield of 4HPA to ∼60−70% compared to the ∼100% yield of the wildtype enzyme. As 4-APA has an amino group in place of the hydroxyl group of 4-HPA, we thought that production of variants with substitutions at His120 or Ser146, which are residues close to the phenolic binding site, might provide enzymes that are superior to the wild-type enzyme in the hydroxylation of 4-APA. Several His120 and Ser146 variants with substitutions with negatively charged (H120D and H120E), positively charged (H120K), and neutral (H120N, S146C, S146A) residues were explored for their ability to perform 4-APA hydroxylation. Single turnover reactions of these His120 and Ser146 variants with 4-APA at pH 7.0 were carried out, and the results are shown in Table 1. The H120N, H120D, H120E, and H120K variants could not hydroxylate 4-APA, while S146C was able to hydroxylate 4-APA with a low product yield of 9 ± 3% hydroxylation. Among all of the variants explored, S146A was the most effective enzyme in catalyzing hydroxylation, resulting in a yield (∼95 ± 3% yield) greater than that of the wild-type C

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX

Articles

ACS Chemical Biology

Scheme 1. (a) Reaction of Wild-Type C2 or S146A with Oxygen in the Absence of 4-APA at pH 6.0 or 9.0, (b) Reaction of WildType C2 with Oxygen in the Presence of 4-APA at pH 6.0 or the Reaction of S146A with Oxygen in the Presence of 4-APA at pH 9.0,a (c) Reaction of S146A with Oxygen in the Presence of 4-APA at pH 6.0b

a The rate constant of step V is thought to be faster than the hydroxylation rate (kOH) because the last phase of the stopped-flow results (flavin oxidation) occurred concomitantly with product formation (hydroxylation, rapid-quench flow results; Figure 2a and Supporting Figure S12a). bRate constants are described as kOOH = oxygen reaction, kH2O2 = H2O2 elimination, and kOH = hydroxylation.

C. Reaction of Wild-Type C2:FMNH− with Oxygen in the Presence of 4-APA at pH 6.0 Observed by Stopped-Flow Spectrophotometry. In contrast to the reaction of the wildtype C2 and 4-APA at pH 9.0, the presence of 4-APA at pH 6.0 increases the observed rate constant of the second phase of the reaction to be much faster than the reaction performed in the absence of 4-APA (kreox of 0.039 ± 0.001 s−1 versus ∼0.0018 s−1, Figure 2a, black and blue kinetic traces). To address whether 4-APA can bind to a binary complex of C2:FMNH− before the enzyme reacts with oxygen, kinetic traces of the reaction of C2:FMNH− (syringe A) mixed with an oxygenated solution of 4-APA (syringe B) in the stopped-flow instrument were compared to the reaction of premixed C2:FMNH− and 4APA (syringe A) with oxygenated buffer (syringe B). The results show that 4-APA does not affect formation of the C4ahydroperoxy-FMN even at 10 mM 4-APA (data not shown), implying that the ligand binds to the wild-type enzyme only after the intermediate is formed (Scheme 1b, step II). As the wild-type C2 can hydroxylate 4-APA (2 mM) at pH 6.0 with ∼83 ± 0% yield (Table 2), the data indicate that the second phase observed in Figure 2a is comprised of multiple steps including binding of 4-APA to C4a-hydroperoxyflavin, the

Table 2) agree well with the results reported in previous studies.28,30 B. Reaction of Wild-Type C2:FMNH− with Oxygen in the Presence of 4-APA at pH 9.0 Observed by Stopped-Flow Spectrophotometry. As the reaction of the wild-type C2 with 4-APA at pH 9.0 did not give any hydroxylation product (Table 1), we performed stopped-flow experiments similar to the experiments in section A, but in the presence of 4-APA (2−10 mM) to investigate whether 4-APA can bind to the wild-type enzyme at this pH. Kinetic traces of the reaction monitored at 390 and 446 nm shown in Supporting Figure S11a (black traces) indicate that the reactions of wild-type enzymes in the presence of 4-APA even at 10 mM are very similar to the reaction in the absence of 4-APA (blue traces). These data suggest that at pH 9.0, 4-APA has essentially no effect on the reaction of the reduced wild-type enzyme with oxygen, implying that it cannot bind to the wild-type enzyme. Therefore, the lack of 4-APA hydroxylation at this pH is likely due to the lack of 4-APA binding or binding of 4-APA in an unproductive manner, which does not disrupt the kinetics of C4a-hydroperoxyflavin formation and decay. D

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX

Articles

ACS Chemical Biology

Figure 3. (a) Transient kinetics of the reaction of S146A with oxygen in the presence of 4-APA at pH 6.0 (black kinetic traces). A solution of S146A (110 μM):FMNH− (50 μM) premixed with 4-APA (2 mM) was mixed with solutions of 0.26, 0.62, 1.22, and 2.06 mM oxygen and 4-APA 2 mM in 100 mM NaH2PO4 at pH 6.0. Concentrations are as before mixing. Reactions were monitored at 390 and 446 nm using the stopped-flow spectrophotometer. A plot of peak areas (DAD 220 nm) of the hydroxylated product 3-OH-4-APA obtained from the rapidquench flow experiments under the same conditions is shown by the red filled circles. A single exponential fit of the data is shown by the red line, consistent with an observed hydroxylation rate constant (kOH) of 2.5 ± 0.2 s−1. For reference, kinetics of the same reaction without 4APA at pH 6.0 are shown in blue. (b) A plot of the observed rate constants of the first phase (kobs1) versus oxygen concentrations. A bimolecular constant (kOOH) from this plot was determined as 2.9 × 105 M−1 s−1.

Figure 2. (a) Transient kinetics of the reaction of wild-type C2 with oxygen in the presence of 4-APA at pH 6.0. A solution of wild-type C2 (110 μM):FMNH− (50 μM) premixed with 4-APA (2 mM) was mixed with solutions of 0.26, 0.62, 1.22, and 2.06 mM oxygen and 4APA 2 mM (before-mixing concentrations) in 100 mM NaH2PO4 at pH 6.0 at 4 °C (black kinetic traces). Reactions were monitored at 390 and 446 nm using a stopped-flow spectrophotometer. A plot of peak areas (SIM-168, TIC) of the hydroxylated product 3-OH-4-APA obtained from the rapid-quench flow experiments under the same conditions is shown in red filled circles. A single exponential fit of the data is shown by the red solid line, consistent with an observed hydroxylation rate constant (kOH) of 0.034 s−1. For reference, kinetics of the same reaction without 4-APA at pH 6.0 are shown by the blue traces. (b) A plot of the observed rate constants of the first phase (kobs1) versus oxygen concentration. A bimolecular constant (kOOH) from this plot was determined as 1.1 × 106 M−1 s−1.

first kinetic phase (black versus blue traces, Supporting Figure S12a, Table 2). These results indicate that 4-APA can bind to S146A only after C4a-hydroperoxyflavin formation (Scheme 1b, step II). The second observed phase was comprised of multiple steps, including the binding of 4-APA to C4a-hydroperoxyflavin, the uncoupling reaction to eliminate H2O2 with kH2O2 = 0.24 s−1 (Table 2), the coupling reaction to form the hydroxylated product with kOH = 1.07 s−1 (Table 2), and the dehydration reaction to yield the oxidized enzyme (Scheme 1b, steps II, III, IV, and V). F. Reaction of S146A:FMNH− with O2 in the Presence of 4APA at pH 6.0 Monitored by Stopped-Flow Spectrophotometry. The kinetics of the reaction of S146A:FMNH− with O2 in the presence of 4-APA at pH 6.0 resulted in four phases (Figure 3a, black traces). Unlike the reaction of S146A at pH 9.0, the first and second phases are dependent on oxygen, implying that these two phases are reactions of enzyme-bound reduced flavin with oxygen to form C4a-hydroperoxy-FMN. The first phase occurred during dead-time0.01 s. During this phase, the absorption at 390 nm increased from ∼0.04−0.05 AU without any significant change in absorption at 446 nm. A plot of the observed rate constants of this phase (kobs1) versus oxygen concentration (Figure 3b) yielded a calculated bimolecular rate constant (kOOH) of 2.9 × 105 M−1 s−1. This value is similar to the bimolecular rate constant of the reaction without 4-APA (2.8 × 105 M−1 s−1; Figure 3a, blue traces; Table 2), indicating that this phase is the reaction of S146A:FMNH− with oxygen to form the S146A:C4a-hydroperoxy-FMN intermediate (step I in Scheme 1c). The second

uncoupling reaction to eliminate H2O2, the coupling reaction to form the hydroxylated product, and dehydration to form the oxidized enzyme (Scheme 1b, steps II, III, IV, and V). D. Measurement of the Hydroxylation Rate Constant of the Reaction of Wild-Type C2 and 4-APA at pH 6.0 Using Rapid-Quench Flow Techniques. In order to directly measure the rate constant associated with the hydroxylation of 4-APA, rapid-quench experiments were carried out to measure the amount of 3-OH-4-APA product formed over time. The observed rate constant of hydroxylation (kOH) was determined as ∼0.034 ± 0.004 s−1 (Figure 2a, red filled circles and red line). Because the reaction undergoes bifurcation pathways of coupling and uncoupling reactions (Scheme 1b), the observed hydroxylation rate constant is the summation of the rate constants from both pathways as represented by eq 1. Based on a product ratio of 83% (Table 2), the actual rate constant of hydroxylation (kOH) and H2O2 elimination (kH2O2) under these conditions can be calculated as 0.028 s−1 and 0.006 s−1, respectively (Table 2). ratio of product =

k OH k OH = kobs k OH + k H2O2

(1)

E. Reaction of S146A:FMNH− with O2 in the Presence of 4APA at pH 9.0 Observed by Stopped-Flow Spectrophotometry. Unlike wild-type C2 at pH 9.0 (section B), S146A can bind 4-APA at this pH. The presence of 4-APA increased kreox of the second phase of the S146A reaction with oxygen from 0.028 ± 0.001 s−1 to ∼1.64 ± 0.01 s−1 but did not affect the E

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX

Articles

ACS Chemical Biology

Table 2. Kinetic Rate Constants of Reactions of Wild-Type and S146A at pH 6.0 and 9.0 According the Scheme 1a−c reaction

substrate

pH

% yield of 3-OH-4-APAa

wild-type C2

4-APA

6 9 6 9 6 9 6 9

83 ± 1

wild-type C2 S146A S146A

4-APA

kOOHb (M−1 s−1) 1.1 1.5 1.1 1.5 2.9 2.5 2.9 2.1

100 ± 1 82 ± 3

× × × × × × × ×

106 106 106 106 105 105 105 105

kobsc (s−1)

kH2O2d (× 10−3 s−1) 6 155 (±1) 1.8 (±0.1) 159 (±1)

26.2, 32.2, 38.6, 41.4 240 6.8 (±0.2) 28 (±1)

kOHe (× 10−3 s−1) 28

2500 (±200) 1070

a

%Yield of product with 2 mM 4-APA substrate was calculated based on HPLC chromatograms of the formed product and values from a standard 3OH-4-APA. bkOOH = bimolecular rate constant of C2:FMNH− reacting with oxygen (step I in Scheme 1a,b,c) ckobs2 = observed rate constants of S146A:FMNH−:4APA reacting with 0.13, 0.31, 0.61, and 1.03 mM oxygen at pH 6.0 were determined as 26.2, 32.2, 38.6, and 41.4 s−1, respectively (Scheme 1c, step III). dkH2O2 = rate constant of hydrogen peroxide elimination (Scheme 1a, step II; Scheme 1b, step III) ekOH = 4-APA hydroxylation rate constant (Scheme 1b, step IV; Scheme 1c, step V)

Table 3. Comparison of Mode of Binding and Hydroxylation Rate Constant (kOH) of Wild-Type and S146A Reactions with 4HPA and 4-APA at pH 6.0 wild-type C2 substrate 4-HPA 4-APA

mode of binding −

binds to C2:FMNH and C4a-hydroperoxyFMN only binds to C4a-hydroperoxyFMN

S146A −1

kOH (s ) a

kOH (s−1)

mode of binding

17 (pH 6−10) 0.028 (pH 6.0)



binds to C2:FMNH and C4a-hydroperoxyFMN binds to C2:FMNH− and C4a-hydroperoxyFMN

b

2.6 (pH 7, 9)c 2.5 (pH 6.0)

a Data are from Ruangchan et al.28 bData are from Tongsook et al.30 cThe hydroxylation rate constant and product yield of wild-type C2 with 4-HPA are independent of pH (pH 6.0−10.0).28 As the product yield of S146A and 4-HPA is also constant throughout a pH range of 6.0−9.030 and the hydroxylation rate constants at pH 7.0 and 9.0 are the same, it is likely that the hydroxylation rate constant of S146A and 4-HPA at pH 6.0 is also 2.6 s−1.

independent of oxygen concentration. These phases involve hydroxylation and the dehydration of C4a-hydroxyFMN to return to the oxidized FMN (Scheme 1c, steps V and VI). G. Measurement of the Hydroxylation Rate Constant of the Reaction of S146A and 4-APA at pH 6.0 Using RapidQuench Flow Techniques. The observed hydroxylation rate constant kOH measured using rapid-quench flow techniques (Figure 3a, red filled circles and red line) indicates that the hydroxylation step occurs during ∼0.05−2 s, essentially concomitant with the third phase of the stopped-flow traces (Figure 3a, black traces), which corresponds to a hydroxylation rate constant of 2.5 ± 0.2 s−1 (Figure 3a, red filled circles and red line; Scheme 1c, step V). Because the reaction under these conditions yields ∼100% of product (Table 2), the observed rate constant is thus equal to the actual hydroxylation rate constant (Table 2; Scheme 1c, step V). Interestingly, this hydroxylation rate constant is similar to the hydroxylation rate constant of the reaction of S146A and 4-HPA (Table 3).30 Catalytic Feature that Enables S146A to be Efficient in Hydroxylation of 4-APA. All of the transient kinetic results indicate that the wild-type C2 does not react with 4-APA efficiently (Table 3). In contrast, kinetics of the reactions of S146A with 4-APA and with 4-HPA are similar. Notably, S146A catalyzes hydroxylation of 4-APA and 4-HPA with similar rate constants at pH 6.0 (2.5 s−1 versus 2.6 s−1, Table 3). These results indicate that the interaction between Ser146 and the phenolic group of 4-HPA is a key dictating factor in optimizing C2 reactivity toward 4-HPA. Upon removal of the Ser146 hydroxyl group, the enzyme (S146A) loses its ability to discriminate between 4-APA and 4-HPA and can subsequently catalyze hydroxylation of the unnatural substrate 4-APA with similar reactivity as the natural substrate 4-HPA. Density Functional Theory Analysis Explains the pH Effect on the Reaction of Wild-Type C2 and 4-APA.

phase occurred during 0.01−0.1 s and was oxygen dependent, giving an absorption increase at 390 nm of ∼0.04−0.09 AU with no significant change in absorbance at 446 nm. The observed rate constants (kobs2) of this phase in the presence of 0.13, 0.31, 0.61, and 1.03 mM oxygen were determined as 26.2, 32.2, 38.6, and 41.4 s−1, respectively (Table 2). As at pH 6.0, a ternary complex of S146A:FMNH−:4-APA can be formed before reaction with oxygen (Supporting Information Figure S13a,b), the second phase in Figure 3a is likely a bimolecular reaction of the S146A:FMNH−:4-APA complex with oxygen to form the S146A:C 4a -hydroperoxy-FMN:4-APA complex (Scheme 1c, step III). In principle, observed rate constants of this phase should be linearly dependent on oxygen concentration. However, as their observed rate constant values are not well separated from the first phase of reaction, accurate observed rate constants could not be obtained. Therefore, a bimolecular rate constant (kOOH) for the second phase cannot be determined. This reaction is similar to the reaction of wildtype C2 with 4-HPA,26,28 in that it has two pathways for oxygen reaction. In the first pathway, S146A:FMNH− can react with oxygen directly to form S146A:C4a-hydroperoxyFMN (with a bimolecular rate constant, kOOH, of 2.9 × 105 M−1 s−1) before binding 4-APA to form the S146A:C4a-hydroperoxyFMN:4APA complex (Scheme 1c, steps I and IV). In the second pathway, S146A:FMNH− can prebind with 4-APA to form S146A:FMNH−:4-APA and then react with oxygen to generate the ternary complex (Scheme 1c, steps II and III). The third phase occurred at 0.1−1.2 s with an absorption decrease at 390 nm (∼0.015−0.025 AU), corresponding to an observed rate constant (kobs3) of ∼1.5 ± 0.2 s−1. The last phase occurred at 1.2−80 s with a large absorption increase at 446 nm (∼0.28 AU) concomitant with an absorption decrease at 390 nm. The observed rate constant of this phase (kobs4 or kreox) was determined as ∼0.048 ± 0.001 s−1. The last two phases were F

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology



ACKNOWLEDGMENTS This research was financially supported by The Thailand Research Fund through grants RTA5980001 (to P.C.) and Royal Golden Jubilee Ph.D. Program Grant PHD/0335/2551 (to T.D.) and RSA5980028 (to P.S.), a grant from Mahidol University (to P.C. T.D., C.P. and C.T.) and the Center of Excellence for Innovation in Chemistry (PERCH-CIC) (to P.S. and S.V.).

Results in Table 1 and Figure 2 and 3 indicate a higher efficiency in hydroxylation and binding for the wild-type and S146A enzymes at lower pH values. As this effect may be due to the protonation states of 4-APA (−NH2 moiety) and active site residues (such as His120), density functional theory (DFT) was employed to investigate the binding energy of 4-APA to the active site model in the form of C4a-hydroperoxyflavin35 under three states: (A) protonated 4-APA and neutral His120 (Figure 1b), (B) neutral 4-APA and protonated His120 (Figure 1c), and (C) neutral 4-APA and neutral His120 (Supporting Information Figure S14). The DFT results indicate that states A and B which contain an extra proton on 4-APA or His120 have binding energies of −54.3 kcal/mol and −47.3 kcal/mol, respectively. These values are greater than those of state C (−37.2 or −39.8 kcal/mol) in which both moieties are neutral. The greater stabilization in states A and B is mainly due to the presence of stronger hydrogen bond interactions than in state C (Figure S14). States A and B likely represent the protonation state of the C2:4-APA complex at pH 6.0, while state C likely represents the protonation state at pH 9.0. Therefore, the DFT results suggest that at lower pH values, the binding energy of 4-APA is more favorable, allowing the reaction to proceed with higher efficiency under these conditions. Conclusion. In conclusion, our work has demonstrated that it is possible to expand the reactivity of a two-component flavin-monooxygenase to go beyond the natural substrate. The variant S146A can catalyze the hydroxylation of an aniline substrate with similar reactivity as the natural phenolic substrate. The knowledge acquired by these studies may be applied for future enzyme engineering of C2 homologues (implications proposed in the Supporting Information).

■ ■



REFERENCES

(1) Huijbers, M. M., Montersino, S., Westphal, A. H., Tischler, D., and van Berkel, W. J. (2014) Flavin dependent monooxygenases. Arch. Biochem. Biophys. 544, 2−17. (2) Torres Pazmino, D. E., Winkler, M., Glieder, A., and Fraaije, M. W. (2010) Monooxygenases as biocatalysts: classification, mechanistic aspects and biotechnological applications. J. Biotechnol. 146, 9−24. (3) Leisch, H., Morley, K., and Lau, P. C. (2011) Baeyer-Villiger monooxygenases: more than just green chemistry. Chem. Rev. 111, 4165−4222. (4) Torres Pazmino, D. E., Dudek, H. M., and Fraaije, M. W. (2010) Baeyer-Villiger monooxygenases: recent advances and future challenges. Curr. Opin. Chem. Biol. 14, 138−144. (5) Montersino, S., Tischler, D., Gassner, G. T., and van Berkel, W. J. (2011) Catalytic and structural features of flavoprotein hydroxylases and epoxidases. Adv. Synth. Catal. 353, 2301−2319. (6) Toogood, H. S., and Scrutton, N. S. (2013) Enzyme engineering toolbox−a ‘catalyst’for change. Catal. Sci. Technol. 3, 2182−2194. (7) Bornscheuer, U. T., Huisman, G. W., Kazlauskas, R. J., Lutz, S., Moore, J. C., and Robins, K. (2012) Engineering the third wave of biocatalysis. Nature 485, 185−194. (8) Bornscheuer, U. T. (2013) From commercial enzymes to biocatalysts designed by protein engineering. Synlett 24, 150−156. (9) Fagan, R. L., and Palfey, B. A. (2010) Flavin-Dependent Enzymes, in Comprehensive Natural Products II (Liu, H.-W., and Mander, L., Ed.), pp 37−113, Elsevier, Oxford. (10) Chaiyen, P. (2010) Flavoenzymes catalyzing oxidative aromatic ring-cleavage reactions. Arch. Biochem. Biophys. 493, 62−70. (11) Liu, X., Dong, Y., Li, X., Ren, Y., Li, Y., Wang, W., Wang, L., and Feng, L. (2010) Characterization of the anthranilate degradation pathway in Geobacillus thermodenitrif icans NG80−2. Microbiology 156, 589−595. (12) Mizutani, Y., Narikawa, T., Satoh, T., Sakurai, N., Kaji, H., Yamada, S., and Samejima, T. (1999) A New UV method for serum γglutamyltransferase assay using recombinant 4-aminobenzoate hydroxylase as a coupling enzyme. J. Biochem. 126, 347−353. (13) Tsuji, H., Oka, T., Kimoto, M., Hong, Y. M., Natori, Y., and Ogawa, T. (1996) Cloning and sequencing of cDNA encoding 4aminobenzoate hydroxylase form Agaricu bisporus. Biochim. Biophys. Acta, Gene Struct. Expression 1309, 31−36. (14) Bando, N., Ogawa, T., Tsuji, H., and Sasaoka, K. (1990) Enzymatic method for selective determination of 4-aminobenzoic acid in urine. Clin. Chem. 36, 1937−1940. (15) Entsch, B., Massey, V., and Claiborne, A. (1987) paraHydroxybenzoate hydroxylase containing 6-hydroxy-FAD is an effective enzyme with modified reaction mechanisms. J. Biol. Chem. 262, 6060−6068. (16) Gatti, D. L., Entsch, B., Ballou, D. P., and Ludwig, M. L. (1996) pH-dependent structural changes in the active site of p-hydroxybenzoate hydroxylase point to the importance of proton and water movements during catalysis. Biochemistry 35, 567−578. (17) Frank, R., Bahrenberg, G., Christoph, T., Schiene, K., De, V. J., Saunders, D. J., Sundermann, B., and Lee, J. (2008) Novel Vanilloid receptor ligands and the use thereof for the production of pharmaceuticals. PCT Pat. Appl. WO2008125342 A2, Oct 23, 2008. (18) Muro, F., Iimura, S., Yoneda, Y., Chiba, J., Watanabe, T., Setoguchi, M., Takayama, G., Yokoyama, M., Takashi, T., Nakayama, A., and Machinaga, N. (2009) A novel and potent VLA-4 antagonist

METHODS

All methods used in this research are described in the Supporting Information.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acschembio.6b00402. Methods; HPLC-chromatograms of single turnover reactions of 4-APA reactions; identification of product from single turnover reaction of 4-APA by comparison with a standard 3-OH-4-APA; multiple turnover reactions of wild-type C2 and S146A with 4-APA at pH 6.0; 1H and 13C NMR spectra of products from multiple turnover reactions; reactions of wild-type C2 or S146A:FMNH− with oxygen in the absence of substrate at pH 6.0 and pH 9.0 monitored by stopped-flow spectrophotometry; reactions of wild-type C2 and S146A reactions in the presence of 4-APA at pH 9.0 observed by stopped-flow spectrophotometry and rapid-quench technique (PDF)



Articles

AUTHOR INFORMATION

Corresponding Author

*Phone: +66-2201 5596. Fax: +66-2354 7174. E-mail: pimchai. [email protected]. Notes

The authors declare no competing financial interest. G

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX

Articles

ACS Chemical Biology based on trans-4-substituted cyclohexanecarboxylic acid. Bioorg. Med. Chem. 17, 1232−43. (19) Brittain, D. R., Davies, G. M., Johnstone, C., and Large, M. S. (2004) Benzoxazole derivatives for use as integrin inhibitors. PCT Pat. Appl. EP1144393B1, Feb 11, 2004. (20) Fischer, R., Hasegawa, H., Hessler, G., Komura, H., Lehmann, T., Mizoguchi, M., Mueller, G., Okigami, H., Tajimi, M., and Ziegelbauer, K. (2001) Piperidyl carboxylic acids as integrin antagonists. PCT Pat. Appl. WO2001058871A1, Aug 16, 2001. (21) Dunwell, D. W., and Evans, D. (1977) Synthesis and antiinflammatory activity of some 2-aryl-6-benzoxazoleacetic acid derivatives. J. Med. Chem. 20, 797−801. (22) Dawson, W. (1983) Method of treating hypersensitivity disease with benzoxazole derivatives. U.S. Patent No. 4,416,892 A, Nov 22, 1983. (23) Chaiyen, P., Suadee, C., and Wilairat, P. (2001) A novel twoprotein component flavoprotein hydroxylase. Eur. J. Biochem. 268, 5550−5561. (24) Thotsaporn, K., Sucharitakul, J., Wongratana, J., Suadee, C., and Chaiyen, P. (2004) Cloning and expression of p-hydroxyphenylacetate 3-hydroxylase from Acinetobacter baumannii: evidence of the divergence of enzymes in the class of two-protein component aromatic hydroxylases. Biochim. Biophys. Acta, Gene Struct. Expression 1680, 60− 66. (25) Sucharitakul, J., Chaiyen, P., Entsch, B., and Ballou, D. P. (2005) The reductase of p-hydroxyphenylacetate 3-hydroxylase from Acinetobacter baumannii requires p-hydroxyphenylacetate for effective catalysis. Biochemistry 44, 10434−10442. (26) Sucharitakul, J., Chaiyen, P., Entsch, B., and Ballou, D. P. (2006) Kinetic mechanisms of the oxygenase from a two-component enzyme, p-hydroxyphenylacetate 3-hydroxylase from Acinetobacter baumannii. J. Biol. Chem. 281, 17044−17053. (27) Sucharitakul, J., Phongsak, T., Entsch, B., Svasti, J., Chaiyen, P., and Ballou, D. P. (2007) Kinetics of a two-component phydroxyphenylacetate hydroxylase explain how reduced flavin is transferred from the reductase to the oxygenase. Biochemistry 46, 8611−8623. (28) Ruangchan, N., Tongsook, C., Sucharitakul, J., and Chaiyen, P. (2011) pH-dependent studies reveal an efficient hydroxylation mechanism of the oxygenase component of p-hydroxyphenylacetate 3-hydroxylase. J. Biol. Chem. 286, 223−233. (29) Thotsaporn, K., Chenprakhon, P., Sucharitakul, J., Mattevi, A., and Chaiyen, P. (2011) Stabilization of C4a-hydroperoxyflavin in a two-component flavin-dependent monooxygenase is achieved through interactions at flavin N5 and C4a atoms. J. Biol. Chem. 286, 28170− 28180. (30) Tongsook, C., Sucharitakul, J., Thotsaporn, K., and Chaiyen, P. (2011) Interactions with the substrate phenolic group are essential for hydroxylation by the oxygenase component of p-hydroxyphenylacetate 3-hydroxylase. J. Biol. Chem. 286, 44491−44502. (31) Phongsak, T., Sucharitakul, J., Thotsaporn, K., Oonanant, W., Yuvaniyama, J., Svasti, J., Ballou, D. P., and Chaiyen, P. (2012) The Cterminal domain of 4-hydroxyphenylacetate 3-hydroxylase from Acinetobacter baumannii is an autoinhibitory domain. J. Biol. Chem. 287, 26213−26222. (32) Sucharitakul, J., Tinikul, R., and Chaiyen, P. (2014) Mechanisms of reduced flavin transfer in the two-component flavin-dependent monooxygenases. Arch. Biochem. Biophys. 555, 33−46. (33) Alfieri, A., Fersini, F., Ruangchan, N., Prongjit, M., Chaiyen, P., and Mattevi, A. (2007) Structure of the monooxygenase component of a two-component flavoprotein monooxygenase. Proc. Natl. Acad. Sci. U. S. A. 104, 1177−1182. (34) Dhammaraj, T., Phintha, A., Pinthong, C., Medhanavyn, D., Tinikul, R., Chenprakhon, P., Sucharitakul, J., Vardhanabhuti, N., Jiarpinitnun, C., and Chaiyen, P. (2015) p-Hydroxyphenylacetate 3Hydroxylase as a Biocatalyst for the Synthesis of Trihydroxyphenolic Acids. ACS Catal. 5, 4492−4502. (35) Visitsatthawong, S., Chenprakhon, P., Chaiyen, P., and Surawatanawong, P. (2015) Mechanism of Oxygen Activation in a

Flavin-Dependent Monooxygenase: A Nearly Barrierless Formation of C4a-Hydroperoxyflavin via Proton-Coupled Electron Transfer. J. Am. Chem. Soc. 137, 9363−9374.

H

DOI: 10.1021/acschembio.6b00402 ACS Chem. Biol. XXXX, XXX, XXX−XXX